Academic journal article
By Timmer, Marcel P.; Inklaar, Robert; van Ark, Bart
Monthly Labor Review , Vol. 128, No. 7
One of the main features of the resurgence in U.S. labor productivity growth after 1995 is the strong contribution by both wholesale and retail trade. In fact, the productivity performance of these sectors is the foremost reason why the American economy grew so much faster than the European economy over the past decade. (1) Naturally, this has attracted attention to the way in which output and productivity in the trade sector is measured in the United States and Europe. (2)
There is no consensus on how to measure output in retail trade for the purpose of productivity measurement. Many productivity studies, including BLS studies in the Review, use real sales per hour worked as an indicator of labor productivity growth. (3) Using sales volume as an indicator for real trade output assumes that there is a one-to-one relationship between the number of products sold and the trade services delivered. For example, if an automobile dealership sells twice as many cars, it is assumed to deliver twice as many trade services. This assumption may of course be criticized from a statistical viewpoint. For example, with the more intensive use of quality-adjusted price indexes for the deflation of sales values, the resulting sales volume is not such a correct proxy for trade services anymore. Nowadays, this problem is most visible when one measures computer sales. For example, nominal sales of the electronic and appliance stores (NAICS 4431) grew on average at 5 percent per year during 1995-2002. The prices of these products, about half of which are computers, declined on average at an annual rate of 12 percent due to dramatic technical improvements. As a result, sales volume grew by a phenomenal 17 percent annually. But, as pointed out by Jack E. Triplett and Barry R Bosworth: "Electronic stores are in the business of selling boxes that they obtain from the manufacturer ... An index that combines the improvements within the box with changes in the number of boxes bears little relationship to the actual activities of the retail store."4 In the remainder of this article, we call this the "inside-the-box" effect.
An alternative way of measuring retail output, which may circumvent the inside-the-box effect, at least conceptually, is to make a clear distinction between the products sold by a retailer and the retail services delivered. Retailers are seen as supplying services through storing and displaying a selection of goods in convenient locations and making them conveniently available for customers to buy. The goods purchased are not treated as part of the intermediate consumption needed to supply these services, when they are resold with only minimal processing such as grading, cleaning, packaging, and so forth. The difference between the value of the goods sold and the value of the goods that would need to be purchased to replace them is called the margin value. This margin concept of trade output is used in the System of National Accounts which underlies the construction of the national accounts around the world,s
To measure productivity growth, the measures of current margins will need to be converted into volume measures of margin. The first way to do this is by deflating current margins by a margin price index which is directly observed. Recently, BLS has introduced a new initiative to measure margin prices in its Producer Price Index program by surveying the difference between the sales price of a specific item and its acquisition cost. However, so far, these measures cover only a limited number of trade industries and years. (6) The second way is to apply a double deflation technique, that is, to use sales prices and purchase prices to construct an estimated margin price. Indeed if prices of goods purchased and goods sold are measured separately with indexes that use the same techniques for quality adjustment, double deflated measures of the real margin will not suffer from the inside-the-box problem.
However, although most national statistical offices within the Organisation for Economic Co-operation and Development (OECD) use margin-based output in current prices, as yet, few statistical offices actually deflate margin values to derive margin volumes. …